The CHADS 2 score is considered a reliable predictor of stroke/thromboembolism risk in patients with atrial fibrillation (AF). However, thromboembolism can occasionally occur even in patients with AF with low CHADS 2 score (CHADS 2 score = 0 or 1). To investigate the incidence and predictors of left atrial appendage (LAA) thrombus (LAAT) formation in patients with AF, we studied consecutive 543 Japanese patients with AF who underwent transesophageal echocardiography before pulmonary vein isolation from 2008 to 2012. All patients were treated with anticoagulation therapy with warfarin, and their clinical and echocardiographic characteristics were evaluated. LAATs were observed in 35 (6.4%) of 543 patients, and the prevalence was clearly correlated with the patient’s CHADS 2 scores. Of 338 patients with low CHADS 2 score, LAATs were observed in 7 patients (2.1%). By multivariate analysis, increased left atrial volume (≥50 ml), decreased ejection fraction (<56%), and increased brain natriuretic peptide level (>75 pg/ml) were significantly associated with increased prevalence of LAATs, even in patients with low CHADS 2 score. Accordingly, we proposed a new scoring system to predict LAAT (left atrial volume ≥50 ml: score 2; ejection fraction <56%: score 1; brain natriuretic peptide >75 pg/ml: score 1). Patients with a score ≥2 have a greater risk of LAAT, whereas all patients with score ≤1 have no LAATs. Our scoring system is useful for evaluation of the risk of LAAT in patients with AF even with low CHADS 2 score.
Atrial fibrillation (AF) is an important and independent risk factor for thromboembolism, and increases the risk for stroke approximately 5-fold. CHADS 2 score (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus [1 score], stroke/transient ischemic attack [2 score]) has been used as a reliable predictor of stroke risk in such patients. According to the 2014 American Heart Association/American College of Cardiology/Heart Rhythm Society (AHA/ACC/HRS) guideline, patients with a CHADS 2 score ≥2 are considered having greater risk for thromboembolism and are recommended for intensive anticoagulation therapy such as warfarin. However, thromboembolism can occur in 0.5% to 3.3% of subjects, even in patients with AF with low CHADS 2 score (CHADS 2 score = 0 or 1). Moreover, there is no clear agreement on whether transesophageal echocardiogram (TEE) screening should be performed to detect left atrial (LA) or left atrial appendage (LAA) thrombus before pulmonary vein isolation (PVI) in all patients. Thus, the aim of the present study was to investigate the incidence and predictors of LAA thrombus (LAAT) in patients with AF using our retrospective data of 543 patients with AF who underwent TEE examination to screen for the presence of LAAT before PVI.
Methods
From May 2008 to May 2012, 622 consecutive Japanese patients with AF who underwent first session of catheter ablation/PVI at the University of Tsukuba Hospital were confirmed from our database. Of these patients, 79 patients were excluded as follows: 16 without warfarin therapy, 12 with significant mitral valvular disease, 12 with inadequate imaging quality, 6 with unsuccessful TEE, 5 undergoing hemodialysis, and 28 patients in whom CHADS 2 or CHA 2 DS 2 -VASc score could not be evaluated. A final 543 patients were enrolled in the analysis. When the patients underwent repeated TEE during the study period, the first set of TEE data was used for analysis.
The clinical characteristics, echocardiographic data, and laboratory examination data were collected from our medical records at the time of TEE studies. Plasma brain natriuretic peptide (BNP) was measured by fluorescence enzyme immunoassay. CHADS 2 score ranging from 0 to 6 was calculated based on the following criteria: congestive heart failure (CHF) = 1 point; hypertension (HT) = 1 point; age ≥75 years = 1 point; diabetes mellitus (DM) = 1 point; history of stroke, transient ischemic attack (TIA), or systemic embolism = 2 points. CHF was defined on the basis of the original definition of CHADS 2 score as existence of clinical symptom and signs of heart failure within the past 3 months or recorded LV systolic dysfunction. DM was diagnosed on the basis of the clinical criteria from the American Diabetes Association. A history of stroke, TIA, or systemic embolism was defined according to the criteria from American Heart Association. CHA 2 DS 2 -VASc score [CHF = 1 point; HT = 1 point; age ≥75 years = 2 points; DM = 1 point; history of stroke, TIA, or systemic embolism = 2 points; vascular disease = 1 point; age from 65 to 74 = 1 point; female gender = 1 point] was also calculated as it was reported to improve the risk stratification of patients with AF with a CHADS 2 score of 0 to 1. According to previous reports, patients with a CHADS 2 score of 0 or 1 were defined as the low CHADS 2 group, whereas those with a score ≥2 were defined as the high CHADS 2 group. The patients with CHA 2 DS 2 -VASc <2 were defined as the low CHA 2 DS 2 -VASc group because the patients with ≥2 were recommended anticoagulation therapy in AHA/ACC/HRS 2014 guideline. Persistent AF was defined as that sustained for longer than 7 days or that requiring cardioversion.
TEE was performed with an ultrasound system equipped with an X7-2t TEE transducer (Phillips iE33; Phillips Medical Systems, Andover, MA), and transthoracic echocardiography (TTE) was performed with a Vivid 9 cardiovascular ultrasound system (GE Healthcare, Horten, Norway). Within 48 hours before PVI, all patients underwent TEE and TTE and were evaluated whether thrombus was present or not in the LA by TEE. During evaluation, LAA was imaged in multiple views to inspect for the thrombus. LAAT was defined as a mass in the LAA of body of atrium and was distinct from the underlying endocardium and pectinate muscles in the multiple views. Left atrial volume (LAV) was measured by the area–length method in apical 4-chamber view. Left ventricular (LV) end-diastolic volume (EDV) and end-systolic volume (ESV) were measured from the apical 4- and 2-chamber views using the Simpson method. LVEF was calculated by the following formula: (EDV − ESV) × 100/EDV.
As anticoagulation therapy, all patients were routinely treated with warfarin. On the basis of the guideline of Japanese Circulation Society, therapeutic prothrombin time/international normalized ratio (PT-INR) was controlled between 2.0 and 3.0 for patients aged <70 years of age or between 1.6 and 2.6 for those aged ≥70 years. PT-INR was arbitrarily checked after starting anticoagulation. In addition, time in therapeutic range (TTR) was evaluated, as calculated by the ratio of duration within the targeted therapeutic range of PT-INR to total duration with warfarin therapy. The cumulative total of patient treatment days was the denominator, and the total number of days that PT-INR was in the target range was the numerator.
Data are presented as mean ± standard deviation (SD) or number (%). For comparison of categorical variables, the chi-square test or Fisher’s exact test was performed as appropriate. Continuous variables between 2 groups were analyzed by the unpaired Student t test or Mann–Whitney U test. To determine the cut-off value in each parameter, the receiver operating characteristic curves plotting sensitivity against specificity were generated. The best cut-off value was defined as the point with the highest sum of sensitivity and specificity. Logistic regression models were used to evaluate the univariate and multivariate influence of each factor. Multivariate logistic regression was performed to identify the significant predictors of the LAAT other than CHADS 2 or CHA 2 DS 2 -VASc score. We modeled a multivariate analysis that included age, gender, EF, LAV, and BNP that yielded significance in the univariate analysis. The presence of heart disease and persistency of AF were not included into multivariate analysis because they were collinear with EF and LAV, respectively. To evaluate the risk of LAAT, odds ratio (OR) and 95% of confidential interval (CI) of each factor were calculated. A p value <0.05 was considered statistically significant. The relative values of the β coefficients in the independent predictors were used to devise the LAAT risk scoring system. All analyses were performed using SPSS statistical software (Version 21.0; Chicago, Illinois).
Results
The clinical and echocardiographic characteristics of all patients (n = 543) are summarized in Table 1 . One hundred fifty (28%) patients had other heart diseases, including 44 ischemic heart diseases, 33 valvular heart diseases except significant mitral valve dysfunction (including 23 aortic valve regurgitations, 9 aortic valve stenoses, and 1 tricuspid valve regurgitation), 38 cardiomyopathies, 17 sick sinus syndromes, and 18 others. All patients were treated with warfarin, and PT-INR was within the target range in 71% of patients at the time of TEE. The prevalence of CHADS 2 score is shown in Figure 1 . The low CHADS 2 group included 338 patients (62.2%) with CHADS 2 score of 0 or 1, whereas the high CHADS 2 group included 205 patients (37.8%) with CHADS 2 score ≥2. The prevalence of the CHA 2 DS 2 -VASc score is shown in Figure 1 .
| Variable | All Patients (n=543) | Thrombus | P value | |
|---|---|---|---|---|
| Present (n=35) | Absent (n=508) | |||
| Age (years) | 61.9 ± 9.9 | 65.1 ± 8.0 | 61.7 ± 9.9 | 0.018 |
| Male | 440 (81%) | 27 (77%) | 413 (81%) | 0.509 |
| Body Mass Index (kg/m 2 ) | 24.3 ± 3.5 | 23.7 ± 3.4 | 24.3 ± 3.5 | 0.245 |
| Thrombus | 35 (6.4%) | – | – | – |
| CHADS 2 score | 1.3 ± 1.2 | 2.7 ± 1.2 | 1.2 ± 1.1 | <0.001 |
| Low CHADS 2 score | 338 (62%) | 7 (20%) | 331 (65%) | <0.001 |
| CHA 2 DS 2 -VASc score | 2.2 ± 1.6 | 3.9 ± 1.7 | 2.0 ± 1.5 | <0.001 |
| Low CHA 2 DS 2 -VASc score | 226 (42%) | 4 (11%) | 222 (44%) | <0.001 |
| Congestive Heart Failure | 121 (22%) | 22 (63%) | 99 (19%) | <0.001 |
| Hypertension | 300 (55%) | 23 (66%) | 277 (55%) | 0.222 |
| Age ≥75years | 36 (6.6%) | 4 (11%) | 32 (6.3%) | 0.278 |
| Age 65-74years | 202 (37%) | 16 (46%) | 186 (37%) | 0.284 |
| Diabetes Mellitus | 112 (21%) | 12 (34%) | 100 (20%) | 0.051 |
| Stroke/Transient Ischemic Attack | 71 (13%) | 16 (46%) | 55 (11%) | <0.001 |
| Vascular Disease | 157 (29%) | 18 (51%) | 139 (27%) | 0.004 |
| Heart Disease | 150 (28%) | 22 (63%) | 128 (25%) | <0.001 |
| Persistent AF | 206 (38%) | 30 (86%) | 176 (35%) | <0.001 |
| End-Diastolic Volume (ml) | 107.9 ± 35.9 | 143.5 ± 77.0 | 105.4 ± 30.0 | 0.006 |
| Ejection Fraction (%) | 62.9 ± 12.2 | 47.1 ± 17.0 | 64.0 ± 11.0 | <0.001 |
| Left Atrial Volume (ml) | 65.1 ± 28.1 | 98.9 ± 39.7 | 62.7 ± 25.5 | <0.001 |
| Prothrombin Time/International Normalization Ratio | 1.9 ± 0.6 | 1.9 ± 0.6 | 1.9 ± 0.6 | 0.821 |
| Time in the Therapeutic Range (%) | 62.0 ± 34.5 | 61.3 ± 36.3 | 62.1 ± 9.8 | 0.901 |
| Creatinine (mg/dl) | 0.9 ± 0.8 | 1.3 ± 1.7 | 0.9 ± 0.8 | 0.192 |
| Brain Natriuretic Peptide (pg/ml) | 119.0 ± 162.9 | 362.3 ± 333.3 | 103.0 ± 130.0 | <0.001 |
LAATs were observed in 2.1% (7 in 338 cases) in Low CHADS 2 group, 1.8% (4 in 226 cases) in Low CHA 2 DS 2 -VASc group, and 6.4% (35 in 543 cases) of all patients ( Figure 2 ). The comparisons of clinical and echocardiographic characteristics between patients with or without LAAT are summarized in Table 1 . Patients with LAAT were older, more frequently complicated with other heart diseases, greater proportion of persistent AF, and greater CHADS 2 score. In addition, larger LAV, larger EDV, lower LVEF, and higher plasma BNP level were observed compared with those in patients without LAAT. In contrast, PT-INR and TTR were not significantly different between patients with and without LAAT. The prevalence of patients with LAAT for each CHADS 2 score is shown in Figure 1 . The prevalence of LAAT was significantly higher in patients with greater CHADS 2 score (p <0.001). No patients with CHADS 2 score of 0 had LAAT. The prevalence of patients with LAAT for each CHA 2 DS 2 -VASc score is shown in Figure 1 . The prevalence of LAAT was also higher in patients with a greater CHA 2 DS 2 -VASc score (p <0.001).
Multivariate analysis showed that LAV ≥50 ml, LVEF <56%, and BNP >75 pg/ml were significant risk factors of LAAT ( Table 2 ). According to our results, we established a new scoring system for predicting LAAT in patients with AF ( Table 3 ); LAAT predictive score = 2 (LAV ≥50 ml) +1 (EF <56%) +1 (BNP >75 pg/ml). All patients with a score ≤1 had no LAATs. Interestingly, the patients with a score ≥2 had a risk of LAAT, even in the low CHADS 2 and low CHA 2 DS 2 -VASc groups ( Table 3 ). The sensitivity and specificity of the scoring system were 100% and 37%, respectively.
